Letting the Player Fail

January 25, 2019

Back in the golden age of adventure games of that late 1980s and early 1990s, LucasArts and Sierra were the two most prevalent companies in the creation of such games. Their design philosophies differed though in that Sierra games allowed player characters to both die and get into unwinnable situations while LucasArts encouraged exploration by not including dead-end situations or player deaths. Time has proven LucasArts’s design approach to be the stronger as the influence of their games is easily observable in modern adventures while Sierra’s games have not aged well. However, as a fan of both I’ve taken a hard look at their libraries to see what I could learn and adapt into my upcoming Sleuthhounds: Cruise game.

One big issue of many Sierra games was that death was a constant companion. Even simple things like stepping off a pathway could kill your character. In some of the poorer designed situations it was necessary to die in order to even know that something needed to be done in a certain area. To be fair, early LucasArts games, like Maniac Mansion and Zak McKracken and the Alien Mindbenders also allowed you to die, although dangerous situations were much more apparent. Dying was also possible in the later Indiana Jones and the Last Crusade and Indiana Jones and the Fate of Atlantis, but these games also made it apparent when death was imminent. The important difference here with the handling of death between the two companies was the degree of consequence.

For both companies, dying meant that the player would lose any progress they’d made in a game since the last time they’d saved. In a genre that grew during the golden age to be increasingly about story and characters and less about solving collections of random puzzles this was a significant issue. In the few LucasArts games where a player could die it was usually apparent when you were getting into a dangerous situation or about to try something that could be somewhat foolhardy. By contrast, death in Sierra games was often abrupt and felt rather arbitrary. Since you could see death coming in LucasArts games it allowed you to save your game before trying something dangerous. In Sierra games the mantra was always “save early, save often” with the result that you might spend as much time on the save screen as actually playing the game.

As adventure games continued to evolve and move away from the design ideas of Sierra games, death tended to go out of most games. Adventures became safe, which allowed for better narratives that weren’t abruptly ended by a wrong move. In some ways this movement seems to have gone too far in that in general not only is it no longer possible to die but it’s not possible to fail in lesser ways either. For example, be chased by a monster and have it catch up to you and it just sort of flails about until you run away again. Or get caught by pirates while trying to steal their treasure and they just reprimand you and send you on your way, allowing unlimited retries to get the treasure.

If you consider other storytelling mediums, protagonists fail all the time. It’s what helps make them endearing. A protagonist who fails and then has to find a way to overcome that failure storywise is much more interesting than a protagonist who succeeds at everything. Sierra games took the failure thing to extremes with the player losing all progress and the story coming to an abrupt halt when the player dies, but I got to thinking that perhaps a softer type of failure might work really well in my own adventure game.

[Ampson and Homes caught in Tobias Rotterdam's work in progress cabin.]
Ampson and Homes caught in Tobias Rotterdam’s work in progress cabin.

In Sierra games failure basically amounts to death and the game mechanic consequence of losing progress. That’s not a satisfactory way of ending a story. For Cruise I considered that the consequence shouldn’t be game mechanic based in nature but should be narrative based instead. I’ve recently completed a sequence in the game where players have to search the cabin of one Tobias Rotterdam. At the conclusion of the search, the player has a limited number of actions with which to cover any traces of being in the cabin and then to hide in a safe spot so that when Rotterdam enters the Sleuthhounds are not discovered.

To me what makes the sequence interesting is that the player is able to either hide successfully or else can be caught. If the player hide’s then they carry on with their investigation. If the player fails to hide then rather than bringing the game to a halt I use it instead as a storytelling opportunity. Here the consequence is not the loss of play progress but rather an impact on the story.

From a character perspective, clearly Rotterdam is not going to be impressed by having his cabin searched by a couple of detectives. As a result, the player’s reputation with Rotterdam decreases. This may mean that in a future sequence Rotterdam will refuse to help the player and so the player will have to find another path to their goal. In a way the player is actually rewarded for failing as they’ll get an alternate sequence that players who successfully hide from Rotterdam won’t see. There’s also the benefit here that the player learns that their actions or inactions have consequences, which makes decision making more important and hopefully more involving.

My approach to this idea of allowing the player to fail is guided by my experience with storytelling in general. Rather than have it be a complete failure and stated as such, my method has been to have the story carry on. The goal is to tell an interesting story and so my intent is that if the player gets a situation “wrong” then the story just continues. It’s fine to allow the player to fail as long as the narrative is still compelling and ends satisfactorily. A character dying randomly? Not satisfying. A character getting caught and so having to become more resourceful in the future? Now that makes for much better storytelling.